Work with me on AI and data science
Thesis topics for PhD and master's students across AI and data science, from decentralized and trustworthy AI to language models, healthcare and digital twins. I also co-supervise theses with industry partners.
Who I work with
PhD students
Doctoral positions are advertised by the university when funding is available. Watch the News page for openings.
Master's students
Students at Skövde and partner universities looking for a thesis in AI, data science, NLP or decentralized AI.
Industry partners
Companies interested in applied and trustworthy AI, joint projects, or co-supervising a thesis.
PhD students I supervise
Current doctoral students and their thesis topics. I have also supervised more than 50 master's theses, several of which led to publications.
Abed Al Bakkour
University of Skövde, Sweden
Distributed AI in the steel industry.
Distributed AI · Digital twins · Steel manufacturing
Fatiha Ait Baali
Cadi Ayyad University, Morocco
Federated fine-tuning of large language models for domain-specific applications: balancing personalization and privacy.
Federated learning · LLMs · Privacy · Virtual assistants
Chaima Lhasnaoui
Cadi Ayyad University, Morocco
Enabling federated learning on the edge for enhanced privacy in medical imaging (FedMedical).
Federated learning · Medical imaging · Edge AI · IoT
Doctoral examination
Collaborative Predictive Maintenance for Smart Manufacturing: From Wireless Control Systems to Federated Learning
Ali Bemani (PhD) · University of Gävle, Sweden
Open research topics
Starting points for a thesis. Each builds on ongoing research, and you are welcome to propose your own idea in these areas.
Decentralized and federated AI
Training models across organisations and devices without moving the data: robust aggregation, efficient edge learning and hierarchical systems.
Trustworthy and privacy-preserving AI
AI that is robust to attacks, respects privacy and can be explained and audited.
Large language models
Fine-tuning and adapting LLMs for specific domains, federated fine-tuning and retrieval-augmented generation.
Agentic AI
AI agents that plan, use tools and work with people, and how to evaluate them.
Natural language processing
Text understanding, question answering, named entity recognition and conversational AI.
AI for healthcare
Diagnostic and decision-support models, from medical imaging to clinical text, trained without pooling patient data.
Digital twins and industrial AI
Soft sensors, process modelling and distributed learning for manufacturing, such as steel production.
Semantic web and linked data
Knowledge graphs, linked open data, semantic data federation and knowledge-based chatbots.
Data mining and decision analysis
Pattern mining, association rules and multi-criteria decision analysis, building on mcdakit.